Skip to main content
. 2014 Mar 19;348:g1888. doi: 10.1136/bmj.g1888

Table 2.

 Meta-regression of effect modifiers

Variable tested Coefficient (95% CI)l P value Proportion (%) of variation explained by covariate
Agomelatine v placebo (12 studies)
Duration of treatment >6 weeks 0.04947 (−0.2158 to 0.3148) 0.687 −11.20
Low risk of bias 0.1043 (−0.1680 to 0.3766) 0.413 0.35
Allocation concealment 0.1652 (−0.0717 to 0.4021) 0.151 20.38
Agomelatine v antidepressant (13 studies)
Duration of treatment >6 weeks −0.0333 (−0.2557 to 0.1891) 0.748 −15.98
Duration of current episode* 0.0006 (−0.0021 to 0.0033) 0.655 −33.93
Low risk of bias 0.0191 (−0.1886 to 0.2268) 0.843 −17.46
Allocation concealment 0.1645 (−0.0165 to 0.3456) 0.071 38.47

*12 studies.